Supersampling will make absolutely no difference. Turn it off to save render time. There was some bad advice in another thread, implying that supersampling could help improve still photo motion. It only helps with media generated by Vegas itself, like titles.

Rendering using "Best" is very important, as Marc already stated.

However, the usual culprit when getting bad results with moving photos is using "track motion" instead of event "pan/crop" to move the photos.

In general, NEVER use track motion to zoom photos. Why? Because Vegas first downsamples the photo to project resolution and then zooms into the photo if you use track motion. By contrast, if you zoom using the pan/crop control in the event, Vegas starts with the full resolution of the photo, and then zooms into that. The difference between these two things can be ENORMOUS: it is not subtle at all.

It is OK to use track motion to move photos; just don't ever use it to zoom (although moving after downsampling might produce some very subtle differences, I guess).

Also, it is OK to use track motion to zoom into video, but only if that video is the same as the project resolution. However, if you put HD video into an SD project, and then want to zoom into that HD video (something that is actually quite useful to do), then the same caveat applies: only use pan/crop to do the zooming.

Johnmeyer gives some good advice. Another thing to think about, convert your photos to a resolution close to the video resolution you are using before you bring them into Vegas. Why? Vegas is not as good at downsampling as a photo program would be. You are likely to get clearer pics this way. If you plan on zooming, I usually make the photo twice the final resolution. This gives me plenty of room to zoom and not lose any quality. But extremely large photos need to be downsampled before brining into Vegas for quality AND performance improvement.

Dave T2

"Vegas is not as good at downsampling as a photo program would be."

Downsampling is relatively easy and by using the "Best" rendering quality option in Vegas, it will look as good as a photo program.

From the Vegas Knowledgebase:

"Best" uses bicubic scaling with integration, while "Good" uses bilinear without integration. You will have a hard time telling the difference, but must use "Best" if you are using high resolution stills or video that are getting scaled down to the final output size. "Good" may have bad artifacts on the near-horizontal edges, while "Best" will look great.

"Another thing to think about, convert your photos to a resolution close to the video resolution you are using before you bring them into Vegas. Why? Vegas is not as good at downsampling as a photo program would be."

Dave,
Actually, bicubic downsampling in Vegas is comparable, if not a bit sharper than bicubic downsampling in Photoshop. I've run several tests on this. However the downsampling method itself is the most significant factor.

Lanczos is visibly superior to bicubic in most circumstances. For instance, Irfanview is better than Vegas at downsizing stills, hands down.

The sage advice that you and I both have adhered to since our Pinnacle Studio days, that the still image pixel dimensions must be exactly match the screen aspect ratio, is less important now than it was then.

That being said, if one brings in stills that are at exactly the SAR dimensions, and does not change that by any means inside Vegas, then the effects of resampling become a non-issue.

Musicvid and John Cline,

From my eyes, my images look better if I downsample them first. I can't say I have done side by side comparisons because it is hard to compare an interlaced video frame to the original, but when I have let Vegas do the downsampling from a larger image, it seems soft to me. But I trust you both in what you are saying.

Dave T2

@johnmeyer
> "Supersampling will make absolutely no difference. It only helps with media generated by Vegas itself, like titles."

Look at clip. Obviously, the supersampling affects the video. What do you think about it?

Dave,
No resampling scheme will duplicate exact bit-for-bit sizing. That is acknowledged.

In the event that resampling is needed to downsize or upsize, there is at least one method that is better than bicubic in either Vegas or Photoshop.

Lanczos 3 is amazing if one needs to resize. Lanczos 4 is still a WIP, from what little I've read.

@altervic,
I'm not johnmeyer, who is more qualified than me.
Sorry, the blurring makes the face appear "fatter," but not sharper. Mark me unimpressed.

Sorry, the blurring makes the face appear "fatter," but not sharper.

Thereby implying that is does effect the video. The smoother lack of judder is a slap in the face only for the proud.

Folks it’s not about impressing people with fractions and minute details, it’s about sharing,learning and improving our skills and getting the most out of Sony Vegas. Mark me as willing to learn.

Thereby implying that is does effect the video.

No, only implying that the horizontal blurring affects the video, negatively.
As far as any effect of supersampling on external source material, I respectfully defer to johnmeyer, above.

And as far as sharing "fractions and minute details," that is precisely what creates "learning and improving our skills and getting the most out of Sony Vegas."

As far as "impressing people," I doubt very much of that actually occurs on the internet in this day and age . . .

“No, only implying that the horizontal blurring affects the video, negatively”

Thereby implying that IT DOES EFFECT THE VIDEO, negatively, positively, upside down this way that way. It effects it.

What?? Supersampling and Blurring are two different things, entirely.
("Affect" is the correct verb).

I actually did do a test before I posted above, just to make sure. I put the results of rendering with and without supersampling back on the timeline and then did an A/B comparison. There was absolutely no difference.

Also, from the Vegas help file:

"supersampling cannot improve the appearance of existing video."

Since I am an hour later than you, I assume that means I can get some sleep,
thanks John.

O very good misdirection musicvid…pic on the grammar to make the rest of the statement appear false.


Johnmeyer said only use pan/crop to do the zooming.

Zooming INTO something and zooming ONTO something are entirely different
Johnmeyer is telling me that z axis zooms can be handled by the pan crop window…this I have got to see??????

Ooops John I would be very careful when calling other peoples advice as BAD simply because it differs from your advice.

Well I also did a A/B comparison and the results are visible!
So is my example faked, did I do something extra to get the result???

For goodness sake Supersampling cannot improve the appearance of existing video but it can affect the motion is the point being made.

"For goodness sake Supersampling cannot improve the appearance of existing video but it can affect the motion is the point being made"

No it cannot affect the motion.
Here is a simple demo I made ages ago to show what Supersampling does to animations created in Vegas when motion blur is applied




In that video from memory the amount of motion blur was kept constant and the amount of supersampling increased.

Both Vegas and After Effects offer this form of control over motion blur created in the application. Neither of them out of the box use motion tracking to create motion blur, it is simply done by blending frames. When motion is created through animation then tweened frames can be used to make the motion blur look more natural but it will have no direct affect on the exisiting video although it may have a negative side affect when exisiting video is animated as the motion blur vectors could be wrong.

Bob.

Bob I remember that clip you did to explain my last question on this topic a few weeks ago,
I understand the principle of motion blur and tweening.

We are simply looking at it from different perspectives
From yours no because the actual motion has not changed. from mine since motion and tweening are visual yes it has an effect and has been changed
The appearance of the motion has been affected. Your clip reinforces my statement.

Now in your clip the affect is seen on the alpha channel. This does not preclude the fact that it will have the same perceivable impact on all pixels.

Indeed it comes down to a question of semantics. Reason I get so uppity about this topic is it keeps coming up and newcomers get themselves all hot and bothered thinking they've found some new magic soup when they haven't.

In my opinion whoever decided to call it "supersampling" deserves a public flogging because it conjures up an idea that it's doing something that it obviously doesn't. I've forgotten what the same thing is called in AE but I recall it better describes what it does without having any risk of misleading a user.

What's even worse in the last few years there's image enhancements techniques developed that do use profound interframe analysis to enhance images and they arguably are better deserving of a name like "supersampling".

Bob.

In general, NEVER use track motion to zoom photos. Why? Because Vegas first downsamples the photo to project resolution and then zooms into the photo if you use track motion. By contrast, if you zoom using the pan/crop control in the event, Vegas starts with the full resolution of the photo, and then zooms into that. The difference between these two things can be ENORMOUS: it is not subtle at all.

John Meyer, thank you so much for this tip! I was making a 16:9 photo video where the original images were 4320 x 2432 px and I was wondering why the image became blurry when zooming into them. Now I know!

Lou

>In general, NEVER use track motion to zoom photos.

In general, I agree with John, but if you need track motion zooms, which could be the case if you are doing 3d motion, you might try setting the project to 4K 16:9 24p (4096x2304, 23.976 fps). Then you are working at near full resolution of your photos.

/jerry

... if you need track motion zooms, which could be the case if you are doing 3d motion, you might try setting the project to 4K 16:9 24p (4096x2304, 23.976 fps). Then you are working at near full resolution of your photos.I've never tried this, but it should work. Pretty neat trick, actually. The only downside I can think of is that preview speed might be slower, perhaps quite a bit slower because any time the project properties don't match the properties of the source media, preview speed is slower.

I know that some people use project properties that are higher than either source or final render resolution in order to get higher-quality results from media that Vegas generates itself, like titles.

Nice posts, jerry and john.

I didn't believe that trick worked until I actually tried it with generated text; setting properties for oversized stills never occured to me. Will try it out.